Front-loaded usefulness matters more than a universal character target.
All tools
PREVIEW TEXT / SOURCE ORDER / BODY BLEED / QAWrite the snippet.
Write the snippet.
Control the fallback.
Design the intended second line, inspect what the final HTML exposes first, and stress-test body bleed before an inbox chooses the snippet for you.
- Source-order inspection
- Seven width scenarios
- HTML + QA manifest
CONTROL100/100
CONTROLLED
Intentional preview with a testable fallback.
The copy and selected implementation have no blocking mechanical finding. Compile, send, and verify the received message.
First exposed sourceNone foundPaste final HTML to inspect source order.
Early visible, hidden, link, and image-alt candidates.
Directional scenarios that append exposed body content.
Shared meaningful words, not a semantic quality score.
Inbox stress preview
Directional layout, not a client screenshotN
Northstar
Your July campaign plan is ready — Three sends, two audience segments, and the goal for each—ready to review.
Gmail · Desktop inbox uses an approximate stress budget of 92 graphemes here. Actual space changes with UI, sender, subject, font, viewport, settings, and product updates.
Width and fallback matrix
7 scenariosClient surfaceSubjectSnippet outcomeState
Exposed source order
Static HTML approximationNo usable text source found.
Add intentional preview text or meaningful body copy near the start of the compiled email.
Copy and implementation findings
4 checksUsable preview-text shape. The copy is a complete, front-loadable fragment.
Subject and preview are differentiated. Meaningful-word overlap is 25%. Review meaning manually; lexical difference does not guarantee useful context.
No mechanical early-source warning. Final received-client testing is still required.
Hidden HTML requires real-client testing. Hiding rules and entity behavior vary after provider and client processing.
Implementation artifact
Hidden preheader HTMLPlace this block immediately after the opening body tag or at the earliest provider-supported template position:
<div style="display:none!important;mso-hide:all;max-height:0;max-width:0;opacity:0;overflow:hidden;visibility:hidden;font-size:1px;line-height:1px;color:transparent;" aria-hidden="true"> Three sends, two audience segments, and the goal for each—ready to review. <span aria-hidden="true">͏‌ ͏‌ ͏‌ ͏‌ </span> </div>
FINAL-BUILD QA
Keep the intended copy, exposed sources, and stress outcomes together.
The manifest records implementation assumptions and every directional result. It does not replace sending the final compiled message to representative inboxes.
Read the full agent guideTHE EXTRACTION CHAIN
The inbox gets the final build, not the content brief.
Preview text can come from a provider field, a hidden block near the body start, or the earliest extractable message content. Template compilation, CSS inlining, link rewriting, image-alt text, navigation, localization, and client behavior can all change the effective snippet.
This tool makes the source order visible so “View in browser,” a hero image description, or repeated headline copy does not become an accidental second line.
- 01Write
Add a useful fact that continues the subject instead of repeating it.
- 02Implement
Prefer a supported provider field; otherwise place tested hidden HTML first.
- 03Compile
Inspect the final HTML after templates, merge tags, and CSS transformations.
- 04Stress
Review narrow and wide outcomes, body bleed, localization, and missing data.
- 05Send
Verify representative real inboxes and archive the received evidence.
OPERATOR QUESTIONS
What can actually be controlled?
The sender controls source content and implementation. The mailbox controls the final amount, extraction, styling, and whether a snippet appears.
How long should preview text be?+
There is no universal safe length. Put the most useful new information first, write a complete thought, and test the final message. The matrix uses directional stress budgets, not published client limits.
Why does body text appear after my preview?+
A mailbox can append extractable body content when the intentional preview does not fill its available snippet area. A small buffer may reduce this in some clients, but a useful body opening is still the safest fallback.
Should I use invisible filler?+
Use the smallest tested amount only when body bleed is genuinely harmful. Invisible entities are not standardized preview controls and can surface, create odd spacing, affect assistive workflows, or change under provider processing.
Can the scanner perfectly reproduce Gmail or Outlook extraction?+
No. It provides a static source-order approximation and directional width scenarios. Clients use private, changing extraction logic and can consider markup differently.
Why scan image alt text?+
Image alternative text can appear early in extractable message content or become visible when images are blocked. It should be useful and contextual, but the snippet should not depend on an accidental hero description.
Should hidden preview text use aria-hidden?+
The generated block uses aria-hidden to reduce duplicate announcement, but email-client and assistive-technology combinations vary. Test the received message and prefer the provider's supported field when it avoids duplicate body content.